• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.174-174
• /
• 2000

Q-learning, based on discrete state and action space, is a most widely used reinforcement Learning. However, this requires a lot of memory and much time for learning all actions of each state when it is applied to a real mobile robot navigation using continuous state and action space Region-based Q-learning is a reinforcement learning method that estimates action values of real state by using triangular-type action distribution model and relationship with its neighboring state which was defined and learned before. This paper proposes a new Region-based Q-learning which uses a reward assigned only when the agent reached the target, and get out of the Local optimal path with adjustment of random action rate. If this is applied to mobile robot navigation, less memory can be used and robot can move smoothly, and optimal solution can be learned fast. To show the validity of our method, computer simulations are illusrated.

• Computers and Concrete
• /
• 제32권3호
• /
• pp.327-337
• /
• 2023

Concrete is the most widely used building material, with various types including high- and ultra-high-strength, reinforced, normal, and lightweight concretes. However, accurately predicting concrete properties is challenging due to the geotechnical design code's requirement for specific characteristics. To overcome this issue, researchers have turned to new technologies like machine learning to develop proper methodologies for concrete specification. In this study, we propose a highly accurate deep learning-based predictive model to investigate the compressive strength (UCS) of lightweight concrete with natural aggregates (pumice). Our model was implemented on a database containing 249 experimental records and revealed that water, cement, water-cement ratio, fine-coarse aggregate, aggregate substitution rate, fine aggregate replacement, and superplasticizer are the most influential covariates on UCS. To validate our model, we trained and tested it on random subsets of the database, and its performance was evaluated using a confusion matrix and receiver operating characteristic (ROC) overall accuracy. The proposed model was compared with widely known machine learning methods such as MLP, SVM, and DT classifiers to assess its capability. In addition, the model was tested on 25 laboratory UCS tests to evaluate its predictability. Our findings showed that the proposed model achieved the highest accuracy (accuracy=0.97, precision=0.97) and the lowest error rate with a high learning rate (R2=0.914), as confirmed by ROC (AUC=0.971), which is higher than other classifiers. Therefore, the proposed method demonstrates a high level of performance and capability for UCS predictions.

• 한국융합학회논문지
• /
• 제13권5호
• /
• pp.237-247
• /
• 2022

본 연구는 급속한 고령화 시대에서 우리나라의 고령층의 경제활동 수준을 다양한 머신러닝 기법으로 정확히 예측하고자 하였다. 고령층의 경제활동 수준과 기존 연구들은 고령층의 삶의 만족도, 사회보장제도 등과 연관된 인과성 검증을 중심으로 이루어진 데 반해, 본 연구는 다양한 머신러닝 기법으로 고령층의 경제활동 수준을 예측하였으며, 특히 예측조합법을 함께 사용함으로써 예측의 안정성을 도모하였다. 60세 이상의 경제활동참가율, 취업률 등을 종속변수로 하고 가구 특성, 소득, 평균임금 등을 설명변수로 설정하여 서로 다른 특성을 지닌 5가지의 머신러닝 기법과 2가지의 예측조합법을 적용하여 예측결과들을 비교하였다. 분석 결과, 종속변수별, 예측구간별로 예측성능이 높은 머신러닝 기법 및 예측조합법은 상이하였으나, 예측의 안정성 측면에서는 예측조합법이 상대적으로 우수한 것으로 나타났다. 이에 따라, 본 연구는 고령층의 경제활동 수준을 정확히 예측하고 예측의 안정성을 도모하여 정책적 관점에서도 실용성을 제고한다고 볼 수 있다.

• 한국산학기술학회논문지
• /
• 제21권7호
• /
• pp.380-390
• /
• 2020

사이버대학교는 20대 중심의 일반대학교 학생보다 사회적 배경, 경제적 요인, IT 지식 및 활용능력 등. 복잡한 교육환경의 변화 요인으로 신입생들의 중도탈락이 높은 실정이다. 따라서 사이버대학교 학생은 일반대학교와 다른 중도탈락 방지 대책과 개선 방법이 필요하다. 본 연구에서는 A 사이버대학의 2017년 및 2018년 1학기 중도탈락에 영향을 미치는 요인을 추출하고 '의사결정트리모델'을 통하여 중점관리 및 상담기준을 분류하여 주요 요인을 도출하였다. 각 주요 요인에 대하여 의사결정 적용기준과 주차별 추진방법을 제시하여 '중도탈락개선모형'으로 구현하였다. 그리고 2019년 1학기 신입생을 대상으로 실제로 운영되고 있는 사이버대학 강의운영에 적용하였다. 그 결과 '중도탈락개선모형'을 적용한 신입생의 중도탈락률은 4.2% 감소하였고 학업지속비율은 11.4% 증가하였다. 본 연구의 주요한 의미는 설문지 조사와 사이버대학 LMS(Learning Management System) 학습활동 결과를 동시에 적용하여 객관적인 분석을 하였다는 것이다. 그러나 학생 자료에 대한 정량적인 요인분석은 되었지만, 정성적인 요인분석이 반영되지 못하였고 연구의 구조적인 한계점이 있어 후속연구가 필요하다. 본 연구에서 구현된 개선모형은 사이버대학의 중도탈락률 및 학업지속비율 개선에 유효하게 적용될 것으로 기대한다.

• 한국콘텐츠학회논문지
• /
• 제21권8호
• /
• pp.1-9
• /
• 2021

딥 러닝을 활용한 컴퓨터 비전 연구는 여전히 대규모의 학습 데이터와 컴퓨팅 파워가 필수적이며, 최적의 네트워크 구조를 도출하기 위해 많은 시행착오가 수반된다. 본 연구에서는 네트워크 최적화나 데이터를 보강하는 것과 무관하게 데이터 자체의 특성만을 고려한 CR(Confusion Rate)기반의 유사 이미지 분류 성능 향상 기법을 제안한다. 제안 방법은 유사한 이미지 데이터를 정확히 분류하기 위해 CR을 산출하고 이를 손실 함수의 가중치에 반영함으로서 딥 러닝 모델의 성능을 향상시키는 기법을 제안한다. 제안 방법은 네트워크 최적화 결과와 독립적으로 이미지 분류 성능의 향상을 가져올 수 있으며, 클래스 간의 유사성을 고려해 유사도가 높은 이미지 식별에 적합하다. 제안 방법의 평가결과 HanDB에서는 0.22%, Animal-10N에서는 3.38%의 성능향상을 보였다. 제안한 방법은 다양한 Noisy Labeled 데이터를 활용한 인공지능 연구에 기반이 될 것을 기대한다.

• 한국경영과학회지
• /
• 제37권4호
• /
• pp.73-93
• /
• 2012

In recent years, service delivery systems employing a self-service approach have been rapidly spreading. Since a self-service system provides a lower product price, it attracts more customers. However, some system managers are still hesitant to accept a self-service system, because there is no systematic model to predict its performance. Therefore, this research attempts to provide a systematic and quantitative model to predict the performance of a self-service system, focused specifically on a self-service gas station. Under this model, the traditional queuing theory was adopted to describe the general self-service process, but it is also assumed that some changes occur in both the customer arrival rate and the service performance rate. In particular, the price elasticity was introduced to capture the change in the customer arrival rate, and the existence of learning effect and helpers were assumed to design the changed service performance rate. Under these assumptions, a simulation model for a self-service gas station is established, and three performance measurements, such as average number of customers, average waiting time, and Utilization are observed, depending on the changes in price difference and helper-operating time. In this research, the optimal operation strategy for price differentiation and helper-operating time is proposed in accordance with the level of the customer learning rate. Although this research confines the scope of the study to the self-service gas station model, the results of this research can be applied to any type of self-service system.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1998년도 추계종합학술대회 논문집
• /
• pp.1277-1280
• /
• 1998

We propose a structural learning method of MLP classifiers for a given application using PfSGA (parameter-free species genetic algorithm), which is a combining of species genetic algorithm(SGA) and parameter-free genetic algorithm(PfGA). experimental results show that PfSGA can reduce the learing time of SGA and has no influence of parameter values on structural learning. And we also convince that PfSGA is more efficient than the other methods in the aspect of misclassification ratio, learning rate, and complexity of MLP structure.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1990년도 추계학술대회 논문집 학회본부
• /
• pp.465-469
• /
• 1990

This thesis proposes a new scheme for the recognition of presegmented Hangul characters. The proposed approach is rather insensitive to noise and variation by applying 2 dimensional convolution to learning patterns. In this thesis, the hangul recognition neural network is implemented in the basis of this scheme and recognition rate is analyzed in boo cases of learning which are learning by binary patterns and learning by binary patterns and convoluted patterns together.

• Journal of Computing Science and Engineering
• /
• 제5권4호
• /
• pp.305-313
• /
• 2011

Due to the expansion of high-speed Internet access, the need for secure and reliable networks has become more critical. The sophistication of network attacks, as well as their severity, has also increased recently. As such, more and more organizations are becoming vulnerable to attack. The aim of this research is to classify network attacks using neural networks (NN), which leads to a higher detection rate and a lower false alarm rate in a shorter time. This paper focuses on two classification types: a single class (normal, or attack), and a multi class (normal, DoS, PRB, R2L, U2R), where the category of attack is also detected by the NN. Extensive analysis is conducted in order to assess the translation of symbolic data, partitioning of the training data and the complexity of the architecture. This paper investigates two engines; the first engine is the back-propagation neural network intrusion detection system (BPNNIDS) and the second engine is the radial basis function neural network intrusion detection system (BPNNIDS). The two engines proposed in this paper are tested against traditional and other machine learning algorithms using a common dataset: the DARPA 98 KDD99 benchmark dataset from International Knowledge Discovery and Data Mining Tools. BPNNIDS shows a superior response compared to the other techniques reported in literature especially in terms of response time, detection rate and false positive rate.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2019년도 추계학술대회
• /
• pp.180-181
• /
• 2019

This paper proposes a rate control algorithm for intra coding frame in HEVC encoder using a deep learning approach. The proposed algorithm is designed for CTU level bit allocation in intra frame by considering visual features spatially and temporally. Our features are generated using visual geometry group (VGG-16) with deep convolutional layers, then it is used for bit allocation per each CTU within an intra frame. According to our experiments, the proposed algorithm can achieve -2.04% Luma component BD-rate gain with minimal bit accuracy loss against the HM-16.20 rate control model.